How would i linearize a max/min constraint for a quadratic program / linear program? Lets say I have 2 constraints, h_1(x) and h_2(x). If i need to enforce a variable alpha to be greater than max(h_1, h_2), this can be accomplished as 2 linear constraints where alpha >= h_1, alpha >=h_2. (Similar, if alpha <= min(h_1, h_2), this maps to alpha <= h_1, alpha <= h_2) But what if I needed to enforce alpha <= max(h_1, h_2)? [alpha >= min(h_1, h_2) would follow similar approach]